Improve doctype preparse performance by aborting as soon as possible, fix required HTML::Parser version.

5 files changed, 21 insertions, 15 deletions

diff --git a/htdocs/docs/install.html b/htdocs/docs/install.html
index 8416595..9550cb8 100755
--- a/htdocs/docs/install.html
+++ b/htdocs/docs/install.html
@@ -1,5 +1,5 @@
-<!--#set var="revision" value="\$Id: install.html,v 1.43 2009-01-04 16:54:17 ville Exp $"
---><!--#set var="date" value="\$Date: 2009-01-04 16:54:17 $"
+<!--#set var="revision" value="\$Id: install.html,v 1.44 2009-01-07 22:19:31 ville Exp $"
+--><!--#set var="date" value="\$Date: 2009-01-07 22:19:31 $"
 --><!--#set var="title" value="Installation Documentation for The W3C Markup Validation Service"
 --><!--#set var="relroot" value="../"
 --><!--#include virtual="../header.html" -->
@@ -136,7 +136,7 @@ install Bundle::W3C::Validator
         <dd>
           Support for multiple character encodings.
         </dd>
-        <dt><a href="http://search.cpan.org/dist/HTML-Parser/">HTML-Parser</a> &gt;= 3.25</dt>
+        <dt><a href="http://search.cpan.org/dist/HTML-Parser/">HTML-Parser</a> &gt;= 3.24</dt>
         <dd>
           Minimal HTML parser used for preparse and finding metadata.
         </dd>
diff --git a/httpd/cgi-bin/check b/httpd/cgi-bin/check
index 862f484..aedb1f8 100755
--- a/httpd/cgi-bin/check
+++ b/httpd/cgi-bin/check
@@ -14,7 +14,7 @@
 # This source code is available under the license at:
 #     http://www.w3.org/Consortium/Legal/copyright-software
 #
-# $Id: check,v 1.636 2009-01-07 22:04:23 ville Exp $
+# $Id: check,v 1.637 2009-01-07 22:19:30 ville Exp $
 
 #
 # Disable buffering on STDOUT!
@@ -55,7 +55,7 @@ use Encode::HanExtra          qw(); # for some chinese character encodings,
 use Encode::JIS2K             qw(); # ditto extra japanese encodings
 use File::Spec::Functions     qw(catfile);
 use HTML::Encoding       0.52 qw();
-use HTML::Parser         3.25 qw(); # Need 3.25 for $p->ignore_elements.
+use HTML::Parser         3.24 qw(); # Need 3.24 for $p->parse($code_ref)
 use HTML::Template       2.6  qw(); # Need 2.6 for path param, other things.
 use HTTP::Headers::Util       qw();
 use HTTP::Request             qw();
@@ -199,7 +199,7 @@ Directory not readable (permission denied): @_r
 
   #
   # Strings
-  $VERSION =  q$Revision: 1.636 $;
+  $VERSION =  q$Revision: 1.637 $;
   $VERSION =~ s/Revision: ([\d\.]+) /$1/;
 
   #
@@ -2111,8 +2111,7 @@ sub preparse_doctype {
   };
 
   my $start = sub {
-    my $tag  = shift;
-    my $attr = shift;
+    my ($p, $tag, $attr) = @_;
 
     if ($File->{Root}) {
       return unless $tag eq $File->{Root};
@@ -2122,6 +2121,9 @@ sub preparse_doctype {
     if ($attr->{xmlns}) {$File->{Namespace} = $attr->{xmlns}};
     if ($attr->{version}) {$File->{'Root Version'} = $attr->{version}};
     if ($attr->{baseProfile}) {$File->{'Root BaseProfile'} = $attr->{baseProfile}};
+
+    # We're done parsing.
+    $p->eof();
   };
 
   # we use HTML::Parser as pre-parser. May use html5lib or other in the future
@@ -2130,11 +2132,15 @@ sub preparse_doctype {
   # if content-type has shown we should pre-parse with XML mode, use that
   # otherwise (mostly text/html cases) use default mode
   $p->xml_mode(TRUE) if ($File->{Mode} =~ /XML/);
-  $p->ignore_elements('BODY');
-  $p->ignore_elements('body');
   $p->handler(declaration => $dtd, 'text');
-  $p->handler(start => $start, 'tag,attr');
-  $p->parse(join "\n", @{$File->{Content}});
+  $p->handler(start => $start, 'self,tag,attr');
+
+  my $line = 0;
+  my $max = scalar(@{$File->{Content}});
+  $p->parse(sub {
+    return ($line < $max) ? $File->{Content}->[$line++] . "\n" : undef;
+  });
+  $p->eof();
 
   # TODO: These \s here are probably wrong now that the strings are utf8_on
   $File->{DOCTYPE} = '' unless defined $File->{DOCTYPE};
diff --git a/misc/bundle/META.yml b/misc/bundle/META.yml
index 44292e2..da67bd7 100644
--- a/misc/bundle/META.yml
+++ b/misc/bundle/META.yml
@@ -15,7 +15,7 @@ requires:
     Encode::JIS2K:                 0
     File::Spec::Functions:         0
     HTML::Encoding:                0.52
-    HTML::Parser:                  3.25
+    HTML::Parser:                  3.24
     HTML::Template:                2.6
     LWP::UserAgent:                2.032
     Net::IP:                       0
diff --git a/misc/bundle/Makefile.PL b/misc/bundle/Makefile.PL
index 6f2f549..cc04879 100644
--- a/misc/bundle/Makefile.PL
+++ b/misc/bundle/Makefile.PL
@@ -14,7 +14,7 @@ WriteMakefile(
                       Encode::JIS2K         => 0,
                       File::Spec::Functions => 0,
                       HTML::Encoding        => 0.52,
-                      HTML::Parser          => 3.25,
+                      HTML::Parser          => 3.24,
                       HTML::Template        => 2.6,
                       LWP::UserAgent        => 2.032,
                       Net::IP               => 0,
diff --git a/misc/bundle/lib/Bundle/W3C/Validator.pm b/misc/bundle/lib/Bundle/W3C/Validator.pm
index 2800a10..8a12751 100644
--- a/misc/bundle/lib/Bundle/W3C/Validator.pm
+++ b/misc/bundle/lib/Bundle/W3C/Validator.pm
@@ -27,7 +27,7 @@ C<perl -MCPAN -e "install Bundle::W3C::Validator">
  Encode::JIS2K
  File::Spec::Functions
  HTML::Encoding 0.52
- HTML::Parser 3.25
+ HTML::Parser 3.24
  HTML::Template 2.6
  LWP::UserAgent 2.032
  Net::IP